<h1 id="@tsuniya_lili-graphic-madness-and-high-contrast-anime-style-for-anima-(dit)">🩸 @tsuniya_lili — Graphic Madness &amp; High-Contrast Anime Style for Anima (DiT)</h1>A surgical, hyper-compact style LoRA capturing an intense, unhinged, high-contrast aesthetic with expressive ringed eyes, bold lineart, and dramatic cel-shading.---<h3 id="why-this-lora-is-built-different-(3-mb-and-zero-leakage)">🔬 Why this LoRA is built different (3 MB &amp; Zero-Leakage)</h3>Unlike standard brute-force LoRAs that weigh 200MB+ and ruin base anatomy, this model was trained using [<a target="_blank" rel="ugc" href="https://github.com/Disya123/Anima-training-framework">Anima-training-framework</a>] — a bespoke architecture-aware trainer leveraging mechanistic interpretability and Function-Space Prior regularization.- ⚡ Ultra-Compact (~3 MB): Only 1.4M parameters across 5 core DiT blocks (Rank 6 MLP, Rank 2 Modulation, Cross-Attention). Zero bloat.- 🔒 True Zero-Leakage: Without the trigger word, the base model remains 100% untampered <code>preservation_drift &lt; 0.006</code>).- 🎨 Orthogonal Style / Prompt Adherence: The LoRA learns how to render, not what to render. It naturally respects medium modifiers like <code>sketch</code>, <code>monochrome</code>, <code>crosshatching</code>, or full color without fighting your prompt.- 📐 Trained on 1280x1280 native resolution.---<h3 id="recommended-settings">⚙️ Recommended Settings</h3>- Base Model: Anima <code>anima_baseV10</code>)- Trigger Word: <code>manga style, @tsuniya_lili, monochrome</code>- LoRA Weight: <code>0.8 – 1.0</code>- CFG Scale: <code>4.0 – 6.0</code>- Sampler: <code>dpmpp_2m_sde_gpu</code> or <code>euler</code> (20–30 steps)---<h3 id="prompting-tips">💡 Prompting Tips</h3>- Core Style: <code>manga style, @tsuniya_lili, monochrome, crazy smile, ringed eyes, heavy blush, blood splatter, red background, high contrast, dynamic angle</code>- Sketch / Manga Mode: <code>manga style, @tsuniya_lili, monochrome, sketch, rough sketch, crosshatching, spot color</code>- Color Palettes: Works great with limited palettes or bold vibrant backgrounds.---🛠️ Trainer Source Code &amp; Research:Trained with [<a target="_blank" rel="ugc" href="https://github.com/Disya123/Anima-training-framework">Anima Training Framework on GitHub</a>] — an architecture-aware DiT trainer with selective layer-targeting and L2 function-space anchoring.

Files with SHA256: b7f331b9cffd43b3b8ea3be38f374b3a711100c58a5518435e346cd850a713a8